threads / patch / 36628

patch, 2 partsMake it possible to update git_wcwidth()

Subject: [PATCH 2/2] Make it possible to update git_wcwidth()

## tl;dr

4 messages between May 9, 2014 and May 12, 2014. Diffs are folded; open one to read it.

replies: 3people: 3as markdown or json

Torsten Bögershausen· May 9, 2014, 21:51 UTC · lore
The function git_wcwidth() returns for a given unicode code point the
width on the display:
-1 for control characters,
 0 for combining or other non-visible code points
 1 for e.g. ASCII
 2 for double-width code points.

This table had been originally been extracted for one Unicode version, probably 3.2.

Make it possible to update the to a later version of Unicode:

Factor out the table from utf8.c into unicode_width.h and add the script update_unicode.sh to update the table based on the latest Unicode specification files.

Thanks to Peter Krefting <peter@softwolves.pp.se> and Kevin Bracey <kevin@bracey.fi> for helping with their Unicode knowledge

Signed-off-by: Torsten Bögershausen <tboegi@web.de>
---
 .gitignore        |   1 +
 Makefile          |   1 +
 unicode_width.h   | 288 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 update_unicode.sh |  37 +++++++
 utf8.c            |  61 +-----------
 5 files changed, 329 insertions(+), 59 deletions(-)
 create mode 100644 unicode_width.h
 create mode 100755 update_unicode.sh
Show changes to 5 files +329 −59

.gitignore, Makefile, unicode_width.h, update_unicode.sh, utf8.c

diff --git a/.gitignore b/.gitignore
index dc600f9..42294e5 100644
--- a/.gitignore
+++ b/.gitignore
@@ -226,6 +226,7 @@
 /config.mak.autogen
 /config.mak.append
 /configure
+/unicode
 /tags
 /TAGS
 /cscope*
diff --git a/Makefile b/Makefile
index a53f3a8..2e2077a 100644
--- a/Makefile
+++ b/Makefile
@@ -730,6 +730,7 @@ LIB_H += transport.h
 LIB_H += tree-walk.h
 LIB_H += tree.h
 LIB_H += unpack-trees.h
+LIB_H += unicode_width.h
 LIB_H += url.h
 LIB_H += urlmatch.h
 LIB_H += userdiff.h
diff --git a/unicode_width.h b/unicode_width.h
new file mode 100644
index 0000000..4db7803
--- /dev/null
+++ b/unicode_width.h
@@ -0,0 +1,288 @@
+static const struct interval zero_width[] = {
+{ 0x0300, 0x036F },
+{ 0x0483, 0x0489 },
+{ 0x0591, 0x05BD },
+{ 0x05BF, 0x05BF },
+{ 0x05C1, 0x05C2 },
+{ 0x05C4, 0x05C5 },
+{ 0x05C7, 0x05C7 },
+{ 0x0600, 0x0604 },
+{ 0x0610, 0x061A },
+{ 0x061C, 0x061C },
+{ 0x064B, 0x065F },
+{ 0x0670, 0x0670 },
+{ 0x06D6, 0x06DD },
+{ 0x06DF, 0x06E4 },
+{ 0x06E7, 0x06E8 },
+{ 0x06EA, 0x06ED },
+{ 0x070F, 0x070F },
+{ 0x0711, 0x0711 },
+{ 0x0730, 0x074A },
+{ 0x07A6, 0x07B0 },
+{ 0x07EB, 0x07F3 },
+{ 0x0816, 0x0819 },
+{ 0x081B, 0x0823 },
+{ 0x0825, 0x0827 },
+{ 0x0829, 0x082D },
+{ 0x0859, 0x085B },
+{ 0x08E4, 0x08FE },
+{ 0x0900, 0x0902 },
+{ 0x093A, 0x093A },
+{ 0x093C, 0x093C },
+{ 0x0941, 0x0948 },
+{ 0x094D, 0x094D },
+{ 0x0951, 0x0957 },
+{ 0x0962, 0x0963 },
+{ 0x0981, 0x0981 },
+{ 0x09BC, 0x09BC },
+{ 0x09C1, 0x09C4 },
+{ 0x09CD, 0x09CD },
+{ 0x09E2, 0x09E3 },
+{ 0x0A01, 0x0A02 },
+{ 0x0A3C, 0x0A3C },
+{ 0x0A41, 0x0A42 },
+{ 0x0A47, 0x0A48 },
+{ 0x0A4B, 0x0A4D },
+{ 0x0A51, 0x0A51 },
+{ 0x0A70, 0x0A71 },
+{ 0x0A75, 0x0A75 },
+{ 0x0A81, 0x0A82 },
+{ 0x0ABC, 0x0ABC },
+{ 0x0AC1, 0x0AC5 },
+{ 0x0AC7, 0x0AC8 },
+{ 0x0ACD, 0x0ACD },
+{ 0x0AE2, 0x0AE3 },
+{ 0x0B01, 0x0B01 },
+{ 0x0B3C, 0x0B3C },
+{ 0x0B3F, 0x0B3F },
+{ 0x0B41, 0x0B44 },
+{ 0x0B4D, 0x0B4D },
+{ 0x0B56, 0x0B56 },
+{ 0x0B62, 0x0B63 },
+{ 0x0B82, 0x0B82 },
+{ 0x0BC0, 0x0BC0 },
+{ 0x0BCD, 0x0BCD },
+{ 0x0C3E, 0x0C40 },
+{ 0x0C46, 0x0C48 },
+{ 0x0C4A, 0x0C4D },
+{ 0x0C55, 0x0C56 },
+{ 0x0C62, 0x0C63 },
+{ 0x0CBC, 0x0CBC },
+{ 0x0CBF, 0x0CBF },
+{ 0x0CC6, 0x0CC6 },
+{ 0x0CCC, 0x0CCD },
+{ 0x0CE2, 0x0CE3 },
+{ 0x0D41, 0x0D44 },
+{ 0x0D4D, 0x0D4D },
+{ 0x0D62, 0x0D63 },
+{ 0x0DCA, 0x0DCA },
+{ 0x0DD2, 0x0DD4 },
+{ 0x0DD6, 0x0DD6 },
+{ 0x0E31, 0x0E31 },
+{ 0x0E34, 0x0E3A },
+{ 0x0E47, 0x0E4E },
+{ 0x0EB1, 0x0EB1 },
+{ 0x0EB4, 0x0EB9 },
+{ 0x0EBB, 0x0EBC },
+{ 0x0EC8, 0x0ECD },
+{ 0x0F18, 0x0F19 },
+{ 0x0F35, 0x0F35 },
+{ 0x0F37, 0x0F37 },
+{ 0x0F39, 0x0F39 },
+{ 0x0F71, 0x0F7E },
+{ 0x0F80, 0x0F84 },
+{ 0x0F86, 0x0F87 },
+{ 0x0F8D, 0x0F97 },
+{ 0x0F99, 0x0FBC },
+{ 0x0FC6, 0x0FC6 },
+{ 0x102D, 0x1030 },
+{ 0x1032, 0x1037 },
+{ 0x1039, 0x103A },
+{ 0x103D, 0x103E },
+{ 0x1058, 0x1059 },
+{ 0x105E, 0x1060 },
+{ 0x1071, 0x1074 },
+{ 0x1082, 0x1082 },
+{ 0x1085, 0x1086 },
+{ 0x108D, 0x108D },
+{ 0x109D, 0x109D },
+{ 0x1160, 0x11FF },
+{ 0x135D, 0x135F },
+{ 0x1712, 0x1714 },
+{ 0x1732, 0x1734 },
+{ 0x1752, 0x1753 },
+{ 0x1772, 0x1773 },
+{ 0x17B4, 0x17B5 },
+{ 0x17B7, 0x17BD },
+{ 0x17C6, 0x17C6 },
+{ 0x17C9, 0x17D3 },
+{ 0x17DD, 0x17DD },
+{ 0x180B, 0x180E },
+{ 0x18A9, 0x18A9 },
+{ 0x1920, 0x1922 },
+{ 0x1927, 0x1928 },
+{ 0x1932, 0x1932 },
+{ 0x1939, 0x193B },
+{ 0x1A17, 0x1A18 },
+{ 0x1A1B, 0x1A1B },
+{ 0x1A56, 0x1A56 },
+{ 0x1A58, 0x1A5E },
+{ 0x1A60, 0x1A60 },
+{ 0x1A62, 0x1A62 },
+{ 0x1A65, 0x1A6C },
+{ 0x1A73, 0x1A7C },
+{ 0x1A7F, 0x1A7F },
+{ 0x1B00, 0x1B03 },
+{ 0x1B34, 0x1B34 },
+{ 0x1B36, 0x1B3A },
+{ 0x1B3C, 0x1B3C },
+{ 0x1B42, 0x1B42 },
+{ 0x1B6B, 0x1B73 },
+{ 0x1B80, 0x1B81 },
+{ 0x1BA2, 0x1BA5 },
+{ 0x1BA8, 0x1BA9 },
+{ 0x1BAB, 0x1BAB },
+{ 0x1BE6, 0x1BE6 },
+{ 0x1BE8, 0x1BE9 },
+{ 0x1BED, 0x1BED },
+{ 0x1BEF, 0x1BF1 },
+{ 0x1C2C, 0x1C33 },
+{ 0x1C36, 0x1C37 },
+{ 0x1CD0, 0x1CD2 },
+{ 0x1CD4, 0x1CE0 },
+{ 0x1CE2, 0x1CE8 },
+{ 0x1CED, 0x1CED },
+{ 0x1CF4, 0x1CF4 },
+{ 0x1DC0, 0x1DE6 },
+{ 0x1DFC, 0x1DFF },
+{ 0x200B, 0x200F },
+{ 0x202A, 0x202E },
+{ 0x2060, 0x2064 },
+{ 0x2066, 0x206F },
+{ 0x20D0, 0x20F0 },
+{ 0x2CEF, 0x2CF1 },
+{ 0x2D7F, 0x2D7F },
+{ 0x2DE0, 0x2DFF },
+{ 0x302A, 0x302D },
+{ 0x3099, 0x309A },
+{ 0xA66F, 0xA672 },
+{ 0xA674, 0xA67D },
+{ 0xA69F, 0xA69F },
+{ 0xA6F0, 0xA6F1 },
+{ 0xA802, 0xA802 },
+{ 0xA806, 0xA806 },
+{ 0xA80B, 0xA80B },
+{ 0xA825, 0xA826 },
+{ 0xA8C4, 0xA8C4 },
+{ 0xA8E0, 0xA8F1 },
+{ 0xA926, 0xA92D },
+{ 0xA947, 0xA951 },
+{ 0xA980, 0xA982 },
+{ 0xA9B3, 0xA9B3 },
+{ 0xA9B6, 0xA9B9 },
+{ 0xA9BC, 0xA9BC },
+{ 0xAA29, 0xAA2E },
+{ 0xAA31, 0xAA32 },
+{ 0xAA35, 0xAA36 },
+{ 0xAA43, 0xAA43 },
+{ 0xAA4C, 0xAA4C },
+{ 0xAAB0, 0xAAB0 },
+{ 0xAAB2, 0xAAB4 },
+{ 0xAAB7, 0xAAB8 },
+{ 0xAABE, 0xAABF },
+{ 0xAAC1, 0xAAC1 },
+{ 0xAAEC, 0xAAED },
+{ 0xAAF6, 0xAAF6 },
+{ 0xABE5, 0xABE5 },
+{ 0xABE8, 0xABE8 },
+{ 0xABED, 0xABED },
+{ 0xFB1E, 0xFB1E },
+{ 0xFE00, 0xFE0F },
+{ 0xFE20, 0xFE26 },
+{ 0xFEFF, 0xFEFF },
+{ 0xFFF9, 0xFFFB },
+{ 0x101FD, 0x101FD },
+{ 0x10A01, 0x10A03 },
+{ 0x10A05, 0x10A06 },
+{ 0x10A0C, 0x10A0F },
+{ 0x10A38, 0x10A3A },
+{ 0x10A3F, 0x10A3F },
+{ 0x11001, 0x11001 },
+{ 0x11038, 0x11046 },
+{ 0x11080, 0x11081 },
+{ 0x110B3, 0x110B6 },
+{ 0x110B9, 0x110BA },
+{ 0x110BD, 0x110BD },
+{ 0x11100, 0x11102 },
+{ 0x11127, 0x1112B },
+{ 0x1112D, 0x11134 },
+{ 0x11180, 0x11181 },
+{ 0x111B6, 0x111BE },
+{ 0x116AB, 0x116AB },
+{ 0x116AD, 0x116AD },
+{ 0x116B0, 0x116B5 },
+{ 0x116B7, 0x116B7 },
+{ 0x16F8F, 0x16F92 },
+{ 0x1D167, 0x1D169 },
+{ 0x1D173, 0x1D182 },
+{ 0x1D185, 0x1D18B },
+{ 0x1D1AA, 0x1D1AD },
+{ 0x1D242, 0x1D244 },
+{ 0xE0001, 0xE0001 },
+{ 0xE0020, 0xE007F },
+{ 0xE0100, 0xE01EF }
+};
+static const struct interval double_width[] = {
+{ /* plane */ 0x0, 0x1C },
+{ /* plane */ 0x1C, 0x21 },
+{ /* plane */ 0x21, 0x22 },
+{ /* plane */ 0x22, 0x23 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ /* plane */ 0x0, 0x0 },
+{ 0x1100, 0x115F },
+{ 0x2329, 0x232A },
+{ 0x2E80, 0x2E99 },
+{ 0x2E9B, 0x2EF3 },
+{ 0x2F00, 0x2FD5 },
+{ 0x2FF0, 0x2FFB },
+{ 0x3000, 0x303E },
+{ 0x3041, 0x3096 },
+{ 0x3099, 0x30FF },
+{ 0x3105, 0x312D },
+{ 0x3131, 0x318E },
+{ 0x3190, 0x31BA },
+{ 0x31C0, 0x31E3 },
+{ 0x31F0, 0x321E },
+{ 0x3220, 0x3247 },
+{ 0x3250, 0x32FE },
+{ 0x3300, 0x4DBF },
+{ 0x4E00, 0xA48C },
+{ 0xA490, 0xA4C6 },
+{ 0xA960, 0xA97C },
+{ 0xAC00, 0xD7A3 },
+{ 0xF900, 0xFAFF },
+{ 0xFE10, 0xFE19 },
+{ 0xFE30, 0xFE52 },
+{ 0xFE54, 0xFE66 },
+{ 0xFE68, 0xFE6B },
+{ 0xFF01, 0xFF60 },
+{ 0xFFE0, 0xFFE6 },
+{ 0x1B000, 0x1B001 },
+{ 0x1F200, 0x1F202 },
+{ 0x1F210, 0x1F23A },
+{ 0x1F240, 0x1F248 },
+{ 0x1F250, 0x1F251 },
+{ 0x20000, 0x2FFFD },
+{ 0x30000, 0x3FFFD }
+};
diff --git a/update_unicode.sh b/update_unicode.sh
new file mode 100755
index 0000000..000b937
--- /dev/null
+++ b/update_unicode.sh
@@ -0,0 +1,37 @@
+#!/bin/sh
+#See http://www.unicode.org/reports/tr44/
+#
+#Me Enclosing_Mark  an enclosing combining mark
+#Mn Nonspacing_Mark a nonspacing combining mark (zero advance width)
+#Cf Format          a format control character
+#
+UNICODEWIDTH_H=../unicode_width.h
+if ! test -d unicode; then
+	mkdir unicode
+fi &&
+( cd unicode &&
+	if ! test -f UnicodeData.txt; then
+		wget http://www.unicode.org/Public/UCD/latest/ucd/UnicodeData.txt
+	fi &&
+	if ! test -f EastAsianWidth.txt; then
+		wget http://www.unicode.org/Public/UCD/latest/ucd/EastAsianWidth.txt
+	fi &&
+	if ! test -d uniset; then
+		git clone https://github.com/depp/uniset.git
+	fi &&
+	(
+		cd uniset &&
+		if ! test -x uniset; then
+			autoreconf -i &&
+			./configure --enable-warnings=-Werror CFLAGS='-O0 -ggdb'
+		fi &&
+		make
+	) &&
+	echo "static const struct interval zero_width[] = {" >$UNICODEWIDTH_H &&
+	UNICODE_DIR=. ./uniset/uniset --32 cat:Me,Mn,Cf + U+1160..U+11FF - U+00AD |
+	grep -v plane >>$UNICODEWIDTH_H &&
+	echo "};" >>$UNICODEWIDTH_H &&
+	echo "static const struct interval double_width[] = {" >>$UNICODEWIDTH_H &&
+	UNICODE_DIR=. ./uniset/uniset --32 eaw:F,W >>$UNICODEWIDTH_H &&
+	echo "};" >>$UNICODEWIDTH_H
+)
diff --git a/utf8.c b/utf8.c
index b5d8136..b30790d 100644
--- a/utf8.c
+++ b/utf8.c
@@ -80,65 +80,8 @@ static int git_wcwidth(ucs_char_t ch)
 {
 	/*
 	 * Sorted list of non-overlapping intervals of non-spacing characters,
-	 * generated by
-	 *   "uniset +cat=Me +cat=Mn +cat=Cf -00AD +1160-11FF +200B c".
 	 */
-	static const struct interval combining[] = {
-		{ 0x0300, 0x036F }, { 0x0483, 0x0489 }, { 0x0591, 0x05BD },
-		{ 0x05BF, 0x05BF }, { 0x05C1, 0x05C2 }, { 0x05C4, 0x05C5 },
-		{ 0x05C7, 0x05C7 }, { 0x0600, 0x0604 }, { 0x0610, 0x061A },
-		{ 0x064B, 0x065F }, { 0x0670, 0x0670 }, { 0x06D6, 0x06E4 },
-		{ 0x06E7, 0x06E8 }, { 0x06EA, 0x06ED }, { 0x070F, 0x070F },
-		{ 0x0711, 0x0711 }, { 0x0730, 0x074A }, { 0x07A6, 0x07B0 },
-		{ 0x0901, 0x0902 }, { 0x093C, 0x093C }, { 0x0941, 0x0948 },
-		{ 0x094D, 0x094D }, { 0x0951, 0x0954 }, { 0x0962, 0x0963 },
-		{ 0x0981, 0x0981 }, { 0x09BC, 0x09BC }, { 0x09C1, 0x09C4 },
-		{ 0x09CD, 0x09CD }, { 0x09E2, 0x09E3 }, { 0x0A01, 0x0A02 },
-		{ 0x0A3C, 0x0A3C }, { 0x0A41, 0x0A42 }, { 0x0A47, 0x0A48 },
-		{ 0x0A4B, 0x0A4D }, { 0x0A70, 0x0A71 }, { 0x0A81, 0x0A82 },
-		{ 0x0ABC, 0x0ABC }, { 0x0AC1, 0x0AC5 }, { 0x0AC7, 0x0AC8 },
-		{ 0x0ACD, 0x0ACD }, { 0x0AE2, 0x0AE3 }, { 0x0B01, 0x0B01 },
-		{ 0x0B3C, 0x0B3C }, { 0x0B3F, 0x0B3F }, { 0x0B41, 0x0B43 },
-		{ 0x0B4D, 0x0B4D }, { 0x0B56, 0x0B56 }, { 0x0B82, 0x0B82 },
-		{ 0x0BC0, 0x0BC0 }, { 0x0BCD, 0x0BCD }, { 0x0C3E, 0x0C40 },
-		{ 0x0C46, 0x0C48 }, { 0x0C4A, 0x0C4D }, { 0x0C55, 0x0C56 },
-		{ 0x0CBC, 0x0CBC }, { 0x0CBF, 0x0CBF }, { 0x0CC6, 0x0CC6 },
-		{ 0x0CCC, 0x0CCD }, { 0x0D41, 0x0D43 }, { 0x0D4D, 0x0D4D },
-		{ 0x0DCA, 0x0DCA }, { 0x0DD2, 0x0DD4 }, { 0x0DD6, 0x0DD6 },
-		{ 0x0E31, 0x0E31 }, { 0x0E34, 0x0E3A }, { 0x0E47, 0x0E4E },
-		{ 0x0EB1, 0x0EB1 }, { 0x0EB4, 0x0EB9 }, { 0x0EBB, 0x0EBC },
-		{ 0x0EC8, 0x0ECD }, { 0x0F18, 0x0F19 }, { 0x0F35, 0x0F35 },
-		{ 0x0F37, 0x0F37 }, { 0x0F39, 0x0F39 }, { 0x0F71, 0x0F7E },
-		{ 0x0F80, 0x0F84 }, { 0x0F86, 0x0F87 }, { 0x0F90, 0x0F97 },
-		{ 0x0F99, 0x0FBC }, { 0x0FC6, 0x0FC6 }, { 0x102D, 0x1030 },
-		{ 0x1032, 0x1032 }, { 0x1036, 0x1037 }, { 0x1039, 0x1039 },
-		{ 0x1058, 0x1059 }, { 0x1160, 0x11FF }, { 0x1712, 0x1714 },
-		{ 0x1732, 0x1734 }, { 0x1752, 0x1753 }, { 0x1772, 0x1773 },
-		{ 0x17B4, 0x17B5 }, { 0x17B7, 0x17BD }, { 0x17C6, 0x17C6 },
-		{ 0x17C9, 0x17D3 }, { 0x17DD, 0x17DD }, { 0x180B, 0x180D },
-		{ 0x18A9, 0x18A9 }, { 0x1920, 0x1922 }, { 0x1927, 0x1928 },
-		{ 0x1932, 0x1932 }, { 0x1939, 0x193B }, { 0x200B, 0x200F },
-		{ 0x202A, 0x202E }, { 0x2060, 0x2063 }, { 0x206A, 0x206F },
-		{ 0x20D0, 0x20EA }, { 0x302A, 0x302F }, { 0x3099, 0x309A },
-		{ 0xFB1E, 0xFB1E }, { 0xFE00, 0xFE0F }, { 0xFE20, 0xFE23 },
-		{ 0xFEFF, 0xFEFF }, { 0xFFF9, 0xFFFB }, { 0x1D167, 0x1D169 },
-		{ 0x1D173, 0x1D182 }, { 0x1D185, 0x1D18B },
-		{ 0x1D1AA, 0x1D1AD }, { 0xE0001, 0xE0001 },
-		{ 0xE0020, 0xE007F }, { 0xE0100, 0xE01EF }
-	};
-	static const struct interval double_width[] = {
-		{ 0x1100, 0x115F },
-		{ 0x2329, 0x232A },
-		{ 0x2E80, 0x303E },
-		{ 0x3040, 0xA4CF },
-		{ 0xAC00, 0xD7A3 },
-		{ 0xF900, 0xFAFF },
-		{ 0xFE30, 0xFE6F },
-		{ 0xFF00, 0xFF60 },
-		{ 0xFFE0, 0xFFE6 },
-		{ 0x20000, 0x2FFFD },
-		{ 0x30000, 0x3FFFD }
-	};
+#include "unicode_width.h"
 
 	/* test for 8-bit control characters */
 	if (ch == 0)
@@ -147,7 +90,7 @@ static int git_wcwidth(ucs_char_t ch)
 		return -1;
 
 	/* binary search in table of non-spacing characters */
-	if (bisearch(ch, combining, sizeof(combining)
+	if (bisearch(ch, zero_width, sizeof(zero_width)
 				/ sizeof(struct interval) - 1))
 		return 0;
 
-- 
1.9.2.691.g8d8dc6d
Peter Krefting· May 12, 2014, 12:38 UTC · re: Torsten Bögershausen · lore

Re: [PATCH 2/2] Make it possible to update git_wcwidth()

Torsten Bögershausen:
Show 6 quoted lines
> The function git_wcwidth() returns for a given unicode code point the
> width on the display:
> -1 for control characters,
> 0 for combining or other non-visible code points
> 1 for e.g. ASCII
> 2 for double-width code points.

This all looks sane, but the problem is that this is also context-dependent since there are a lot of characters with "ambiguous" widths, i.e., characters that are "double" width for CJK locales (and fonts), but "single" width for others. This includes Greek and Cyrillic characters, which are encoded using the double-byte parts of the CJK DBCS encodings.

I'm not quite sure how much impact this would have on day-to-day Git operation in a CJK locale, however, as I guess they would mostly encounter texts in their own language (which would mostly be "double" width) or English (which would be unambiguously "single" width).

Anyone on the list running Git in a CJK locale that would like to weigh in here?

-- 
\\// Peter - http://www.softwolves.pp.se/
Junio C Hamano· May 12, 2014, 17:44 UTC · re: Peter Krefting · lore

Re: [PATCH 2/2] Make it possible to update git_wcwidth()

Peter Krefting <peter@softwolves.pp.se> writes:
Show 23 quoted lines
> Torsten Bögershausen:
>
>> The function git_wcwidth() returns for a given unicode code point the
>> width on the display:
>> -1 for control characters,
>> 0 for combining or other non-visible code points
>> 1 for e.g. ASCII
>> 2 for double-width code points.
>
> This all looks sane, but the problem is that this is also
> context-dependent since there are a lot of characters with "ambiguous"
> widths, i.e., characters that are "double" width for CJK locales (and
> fonts), but "single" width for others. This includes Greek and
> Cyrillic characters, which are encoded using the double-byte parts of
> the CJK DBCS encodings.
>
> I'm not quite sure how much impact this would have on day-to-day Git
> operation in a CJK locale, however, as I guess they would mostly
> encounter texts in their own language (which would mostly be "double"
> width) or English (which would be unambiguously "single" width).
>
> Anyone on the list running Git in a CJK locale that would like to
> weigh in here?

The issue does appear in the real life. A solution I've seen used in a terminaul emulator program was to give the user a choice to say "I want ambiguous ones to be treated as double (or single)". As a J-locale user, I naturally set the configuration to double while using that program (I no longer do).

Junio C Hamano· May 12, 2014, 17:54 UTC · re: Torsten Bögershausen · lore

Re: [PATCH 2/2] Make it possible to update git_wcwidth()

Torsten Bögershausen <tboegi@web.de> writes:
Show 23 quoted lines
> The function git_wcwidth() returns for a given unicode code point the
> width on the display:
> -1 for control characters,
>  0 for combining or other non-visible code points
>  1 for e.g. ASCII
>  2 for double-width code points.
>
> This table had been originally been extracted for one Unicode version,
> probably 3.2.
>
> Make it possible to update the to a later version of Unicode:
>
> Factor out the table from utf8.c into unicode_width.h and
> add the script update_unicode.sh to update the table based on the latest
> Unicode specification files.
>
> Thanks to
> Peter Krefting <peter@softwolves.pp.se> and
> Kevin Bracey <kevin@bracey.fi>
> for helping with their Unicode knowledge
>
> Signed-off-by: Torsten Bögershausen <tboegi@web.de>
> ---
I would say this is a good idea, but a few nitpicks.
Show 6 quoted lines
> diff --git a/unicode_width.h b/unicode_width.h
> new file mode 100644
> index 0000000..4db7803
> --- /dev/null
> +++ b/unicode_width.h
> @@ -0,0 +1,288 @@

Please update the script (and the resulting file) to caution against misuse/mismanagement of this file by adding a comment to at least state:

 - This is a generated file and you are not supposed to modify it; and
 - This is to be included only once from one place in the code and
   that is why this does not use the usual #ifndef X/#define X/#endif
   double-inclusion guards.

An obvious and viable alternative to the second would be to do the usual double-inclusion guard. I do not have much preference either way.

Show 18 quoted lines
> +static const struct interval zero_width[] = {
> ...
> +};
> +static const struct interval double_width[] = {
> ...
> +};
> diff --git a/update_unicode.sh b/update_unicode.sh
> new file mode 100755
> index 0000000..000b937
> --- /dev/null
> +++ b/update_unicode.sh
> @@ -0,0 +1,37 @@
> +#!/bin/sh
> +#See http://www.unicode.org/reports/tr44/
> +#
> +#Me Enclosing_Mark  an enclosing combining mark
> +#Mn Nonspacing_Mark a nonspacing combining mark (zero advance width)
> +#Cf Format          a format control character
Please have a SP after # in these comments to make them readable?
Show 5 quoted lines
> +#
> +UNICODEWIDTH_H=../unicode_width.h
> +if ! test -d unicode; then
> +	mkdir unicode
> +fi &&
Style:
	if ! test -d unicode
        then
        	...
	fi
Show 15 quoted lines
> +( cd unicode &&
> +	if ! test -f UnicodeData.txt; then
> +		wget http://www.unicode.org/Public/UCD/latest/ucd/UnicodeData.txt
> +	fi &&
> +	if ! test -f EastAsianWidth.txt; then
> +		wget http://www.unicode.org/Public/UCD/latest/ucd/EastAsianWidth.txt
> +	fi &&
> +	if ! test -d uniset; then
> +		git clone https://github.com/depp/uniset.git
> +	fi &&
> +	(
> +		cd uniset &&
> +		if ! test -x uniset; then
> +			autoreconf -i &&
> +			./configure --enable-warnings=-Werror CFLAGS='-O0 -ggdb'
What are these "-O0 -ggdb" about?
Show 18 quoted lines
> +		fi &&
> +		make
> +	) &&
> +	echo "static const struct interval zero_width[] = {" >$UNICODEWIDTH_H &&
> +	UNICODE_DIR=. ./uniset/uniset --32 cat:Me,Mn,Cf + U+1160..U+11FF - U+00AD |
> +	grep -v plane >>$UNICODEWIDTH_H &&
> +	echo "};" >>$UNICODEWIDTH_H &&
> +	echo "static const struct interval double_width[] = {" >>$UNICODEWIDTH_H &&
> +	UNICODE_DIR=. ./uniset/uniset --32 eaw:F,W >>$UNICODEWIDTH_H &&
> +	echo "};" >>$UNICODEWIDTH_H
> +)
> @@ -147,7 +90,7 @@ static int git_wcwidth(ucs_char_t ch)
>  		return -1;
>  
>  	/* binary search in table of non-spacing characters */
> -	if (bisearch(ch, combining, sizeof(combining)
> +	if (bisearch(ch, zero_width, sizeof(zero_width)
>  				/ sizeof(struct interval) - 1))

To my eyes, that line looks folded at a funny place. I think it is more conventional to fold after the operator, i.e.

	if (bisearch(ch, zero_width, sizeof(zero_width)	/
				sizeof(struct interval) - 1))
but
	if (bisearch(ch, zero_width,
  		     sizeof(zero_width)	/ sizeof(struct interval) - 1))

may probably be a lot more logical and readable. Maybe it is just me.

← back to recent threads